India's largest AI infrastructure bet. The Mumbai-based company, founded by data center veteran Sharad Sanghi (Netmagic, acquired by NTT for $128M), is building the country's first sovereign AI cloud with plans to deploy 20,000+ GPUs. Blackstone's $1.2B backing at a time when India is aggressively pushing to localize AI compute underscores the geopolitical premium on national GPU capacity. Previously raised $50M across seed and Series A from Z47, Nexus, and NTTVC.
Daily Funding Roundup:
Feb 26, 2026
Neysa lands $1.2B from Blackstone to build India's sovereign AI cloud, headlining a $1.4B+ day. UpGuard resurfaces with $75M after 4+ years. Encord hits $550M valuation for physical AI data infrastructure. Battery Ventures backs Letter AI's $40M Series B for AI sales enablement.
Key Themes
Sovereign AI compute becomes an institutional asset class. Neysa's $1.2B from Blackstone is not a typical VC deal. It is infrastructure capital flowing into national GPU capacity at a time when India is aggressively pushing to localize AI training and inference. The deal structure (equity + debt) mirrors how data centers, not startups, get financed, a signal that AI compute is graduating from venture to institutional real assets.
Physical AI needs its own data stack. Encord's $60M at $550M and RLWRLD's $26M seed extension both attack the same thesis from different angles: robots, drones, and autonomous vehicles need fundamentally different data infrastructure than language models. The convergence of three physical AI deals in a single day suggests this is becoming a consensus bet.
Profitability still wins. UpGuard's 4.5-year gap between funding rounds, during which it grew to 50,000 customers and 100B+ daily risk signals, is the anti-hype story of the day. Springcoast Partners leading a $75M Series C for a company that clearly did not need the money is a vote of confidence in capital efficiency.
The Rounds
The Australian cyber risk posture management company emerged from a 4.5-year funding drought with its biggest round yet. Processing 100+ billion risk signals daily for 50,000 organizations across 90+ countries, UpGuard's growth-without-capital story is a rarity in SaaS. The platform unifies vendor risk, breach risk, and trust exchange in a category (CRPM) that the company largely created. G2 leader in third-party risk management for 15 consecutive quarters.
The YC-backed data infrastructure company for physical AI hit an inflection point: 10x revenue growth from physical AI customers and 5+ petabytes of data on its platform. Encord's thesis is that the next wave of AI (robots, autonomous vehicles, drones) needs fundamentally different data tooling than LLMs. With 300+ AI teams including Toyota Woven, Skydio, and AXA on the platform, the $550M valuation reflects conviction that physical AI data will be as critical as training data was for language models.
The YC S23 startup went from seed to $40M Series B in under two years. Letter AI's platform generates personalized sales enablement and coaching based on live deal context, a step beyond static playbooks. Enterprise customers include Lenovo, Adobe, Novo Nordisk, and Plaid. Battery Ventures leading suggests confidence in revenue metrics. Co-founders Ali Akhtar (ex-McKinsey, project44) and Armen Forget (ex-Samsara) bring a data-first lens to a category historically driven by content management.
Also noted
Acquisitions
Cash-and-stock deal for the ATM and banking technology company. Creates a combined entity managing 1M+ ATMs and self-service kiosks globally. Accelerates Brink's pivot from armored transport to digital banking infrastructure.
News & Signals
Generate Biomedicines files for $400M IPO
The Flagship Pioneering-backed protein design company joins the biotech IPO wave. Uses generative AI to design novel protein therapeutics. Raised $640M+ in private funding. If priced, it would be one of the largest biotech IPOs of 2026.
Open Source Endowment launches with $750K
New initiative backed by Canva, Google, GitHub, and Protocol Labs establishes a permanent fund for critical open source infrastructure. Modeled on university endowments. Addresses the sustainability crisis where projects powering trillion-dollar companies survive on volunteer effort.
Physical AI theme accelerates globally
Three of today's six deals (Encord, RLWRLD, Neysa) directly serve the physical AI stack: data infrastructure, robotics foundation models, and GPU compute. The convergence suggests investors see physical AI as the next major deployment frontier after language models.
VC Mood on X
Bullish signals
- Neysa's $1.2B from Blackstone signals sovereign AI compute becoming a legitimate asset class for institutional capital
- Physical AI convergence: three deals in one day (Encord, RLWRLD, Neysa) validating the post-LLM deployment thesis
- UpGuard's growth without new capital for 4.5 years drew admiration: "This is what real SaaS looks like"
- Letter AI's seed-to-Series B in under two years with Battery leading suggests strong revenue metrics in AI enablement
Bearish signals
- Neysa's deal structure (equity + debt) raised questions about actual equity valuation versus headline number
- Korean robotics investment (RLWRLD) drew skepticism about whether foundation models generalize across industrial environments
- Generate Biomedicines' IPO filing timing prompted debate about whether the biotech window is truly open or just ajar
- Open Source Endowment's $750K launch size described as "a drop in the ocean" relative to the infrastructure it supports
Thursday's mood on X was dominated by two threads: the geopolitics of sovereign AI compute (Neysa) and the emerging physical AI stack. Investors debated whether Blackstone-style infrastructure capital entering AI signals market maturity or peak froth. The Encord discussion centered on whether physical AI data tooling is a durable category or will be subsumed by the robotics companies themselves. UpGuard's comeback story generated the most positive engagement, with operators citing it as proof that profitability still matters. The NODA AI/Bessemer deal quietly drew attention from defense-focused VCs noting the "dual-use AI" thesis is gaining traction at the Series A stage. Overall mood: cautiously constructive, with attention shifting from model companies to infrastructure and deployment plays.
Methodology
Data sourced from company announcements, press coverage, and social media posts via Grok analysis of X. All funding rounds include linked sources in our database. Visit individual company pages to see source URLs. X sentiment is an informal snapshot, not a quantitative index.